Frequency of glucose-6-phosphate dehydrogenase deficiency in malaria patients from six African countries enrolled in two randomized anti-malarial clinical trials

RESEARCH Open Access

Frequency of glucose-6-phosphate dehydrogenasedeficiency in malaria patients from six Africancountries enrolled in two randomized anti-malarialclinical trialsNick Carter1, Allan Pamba2, Stephan Duparc3 and John N Waitumbi4*

Abstract

Background: Glucose-6-phosphate dehydrogenase (G6PD) deficiency is common in populations living in malariaendemic areas. G6PD genotype and phenotype were determined for malaria patients enrolled in thechlorproguanil-dapsone-artesunate (CDA) phase III clinical trial programme.

Methods: Study participants, aged > 1 year, with microscopically confirmed uncomplicated Plasmodium falciparummalaria, and haemoglobin ≥ 70 g/L or haematocrit ≥ 25%, were recruited into two clinical trials conducted in sixAfrican countries (Burkina Faso, Ghana, Kenya, Nigeria, Tanzania, Mali). G6PD genotype of the three most commonAfrican forms, G6PD*B, G6PD*A (A376G), and G6PD*A- (G202A, A542T, G680T and T968C), were determined andused for frequency estimation. G6PD phenotype was assessed qualitatively using the NADPH fluorescence test.Exploratory analyses investigated the effect of G6PD status on baseline haemoglobin concentration, temperature,asexual parasitaemia and anti-malarial efficacy after treatment with CDA 2/2.5/4 mg/kg or chlorproguanil-dapsone2/2.5 mg/kg (both given once daily for three days) or six-dose artemether-lumefantrine.

Results: Of 2264 malaria patients enrolled, 2045 had G6PD genotype available and comprised the primary analysispopulation (1018 males, 1027 females). G6PD deficiency prevalence was 9.0% (184/2045; 7.2% [N = 147] malehemizygous plus 1.8% [N = 37] female homozygous), 13.3% (273/2045) of patients were heterozygous females,77.7% (1588/2045) were G6PD normal. All deficient G6PD*A- genotypes were A376G/G202A. G6PD phenotype wasavailable for 64.5% (1319/2045) of patients: 10.2% (134/1319) were G6PD deficient, 9.6% (127/1319) intermediate,and 80.2% (1058/1319) normal. Phenotype test specificity in detecting hemizygous males was 70.7% (70/99) and48.0% (12/25) for homozygous females. Logistic regression found no significant effect of G6PD genotype onadjusted mean baseline haemoglobin (p = 0.154), adjusted mean baseline temperature (p = 0.9617), or adjustedlog mean baseline parasitaemia (p = 0.365). There was no effect of G6PD genotype (p = 0.490) or phenotype (p =0.391) on the rate of malaria recrudescence, or reinfection (p = 0.134 and p = 0.354, respectively).

Conclusions: G6PD deficiency is common in African patients with malaria and until a reliable and simple G6PDtest is available, the use of 8-aminoquinolines will remain problematic. G6PD status did not impact baselinehaemoglobin, parasitaemia or temperature or the outcomes of anti-malarial therapy.

Trial registration: Clinicaltrials.gov: NCT00344006 and NCT00371735.

* Correspondence: [email protected] Reed Project/Kenya Medical Research Institute, Kisumu, UnitedNations Avenue Gigiri, Village Market, Nairobi 00621, KenyaFull list of author information is available at the end of the article

Carter et al. Malaria Journal 2011, 10:241http://www.malariajournal.com/content/10/1/241

© 2011 Carter et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative CommonsAttribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction inany medium, provided the original work is properly cited.

http://www.clinicaltrials.gov/ct2/show/NCT00344006

http://www.clinicaltrials.gov/ct2/show/NCT00371735

mailto:[email protected]

http://creativecommons.org/licenses/by/2.0

BackgroundGlucose-6-phosphate dehydrogenase (G6PD) is anenzyme in the pentose phosphate pathway (PPP) thatplays an important role in protecting cells from oxida-tive damage by producing NADPH and reduced glu-tathione. In the erythrocyte, which lacks a nucleus,mitochondria and other organelles, PPP is the only bio-chemical pathway for generating reducing capacity [1,2].The G6PD gene is highly polymorphic with almost 400reported variants, conferring varying levels of enzymeactivity [2]. G6PD is X-linked, and so deficient variantsare expressed more commonly in males than in females.Worldwide, an estimated 400 million people are G6PDdeficient with the distribution corresponding to areas inwhich malaria is, or has been, prevalent. This has led tospeculation that malaria has been the selection pressurethat has favoured the maintenance of this potentiallydeleterious trait. There is some evidence that hemizy-gous males may be protected against severe malaria[1,3-5].Although there are over 400 G6PD variants worldwide,

in sub-Saharan Africa three variants occur with poly-morphic frequencies (> 0.1%); G6PD*B, G6PD*A andG6PD*A-. G6PD*B is the wild type and the most commonvariant in Africa and worldwide. G6PD*A has a singleA®G substitution at nucleotide number 376, though is anormal variant with about 90% of the G6PD*B enzymeactivity [6]. G6PD*A- is a deficient variant with about8-20% of the wild type enzyme activity and, in addition tothe A376G mutation that describes the G6PD*A variant,most commonly also involves a G®A substitution atnucleotide 202 [7]. However, G6PD*A- variants with sub-stitutions at 542 G®T, 680 G®T or 968 T®C have alsobeen identified in Africa [3,8].Individuals with G6PD*A- are normally asymptomatic

[1]. However, erythrocyte exposure to oxidative stresscauses haemoglobin denaturation, ultimately resulting inhaemolysis. Haemolytic anaemia in G6PD-deficient indi-viduals can be triggered by a range of oxidative agents,such as infections and certain foods and drugs, includinganti-malarials [9].This study reports the frequency of G6PD genotypes

and phenotypes in malaria patients who participated intwo Phase III clinical trials of chlorproguanil-dapsone-artesunate (CDA) [10,11]. The CDA clinical trial pro-gramme was terminated because of an increased risk ofdrug-related haemolysis with CDA in G6PD-deficientcompared with G6PD-normal malaria patients [10,11].The haemolysis was almost certainly because of the dap-sone component [10-13]. This paper also reports thatG6PD genotype did not influence baseline haemoglobinconcentrations, temperature and asexual parasitaemia orthe outcomes of anti-malarial efficacy.

MethodsThe clinical data from the CDA trials have been pub-lished separately and those reports include more com-prehensive descriptions of the conduct of the individualtrials [10,11].

Ethics statementThis study was conducted in accordance with GoodClinical Practices, applicable regulatory requirements,and the Declaration of Helsinki. Approval was obtainedfrom each participating centre’s ethics committee orinstitutional review board and the WHO Special Pro-gramme for Research and Training in Tropical Diseases(TDR). An Independent Data Monitoring Committee(IDMC) was convened and planned to conduct twointerim safety analyses. The IDMC appointed an inde-pendent end-point reviewer (blinded to treatmentassignment). Written or oral witnessed informed con-sent was required from patients or their parent/guardianplus assent from patients who were ≥ 12-18 years old.

ObjectivesThis study used genotyping and phenotyping to deter-mine the frequency of G6PD deficiency in malariapatients included in the CDA Phase III clinical trial pro-gramme in Africa. Exploratory analyses investigatedwhether baseline haemoglobin concentration, tempera-ture and asexual parasitaemia were influenced by G6PDgenotype. The effect of G6PD status on anti-malarialefficacy was also examined.

ParticipantsPatients were screened for inclusion into two clinicaltrials. Study 005 compared CDA versus artemether-lumefantrine (AL) in malaria patients aged > 1- < 14years of age and was conducted between June 2006 andAugust 2007 at 11 centres in five African countries:Bobo-Dioulasso, Burkina Faso; Kintampo, Ghana;Eldoret, Kilifi and Pingilikani, Kenya; Ibadan, Enugu, Josand Calabar, Nigeria; and Bagamoyo and Kiwangwa, Tan-zania [11]. Study 006 compared CDA versus chlorpro-guanil-dapsone (CPG-DDS) in patients > 1 year old andwas conducted between April 2006 to May 2007 at sevenhealth centres in four African countries: Ouagadougou,Burkina Faso; Kumasi, Ghana; Doneguebougou andBanambani, Mali; and Ile-Ife, Jos and Lagos, Nigeria [10].Across the two trials, eligible subjects included males

and females who met the following inclusion criteria:acute uncomplicated microscopically verified Plasmodiumfalciparum malaria (parasite count 2000 to 200,000 μL-1);fever at enrollment or history of fever; weight ≥ 7.5 kg;haemoglobin ≥ 70 g/L or haematocrit ≥ 25%; willingnessto comply with study procedures.


Page 2 of 14

Subjects were excluded if they had: severe/complicatedP. falciparum malaria; known G6PD deficiency, methae-moglobin reductase deficiency, haemoglobin M/E (i.e.increased risk of methaemoglobinaemia), or porphyria;neonatal hyperbilirubinaemia; received potentially hae-molytic concomitant medication; concomitant infection(including Plasmodium vivax, Plasmodium ovale or Plas-modium malariae); any underlying disease that couldcompromise malaria diagnosis or the anti-malarial effi-cacy assessment; known hypersensitivity/allergy to studytreatments or biguanides, sulphones, sulphonamides, orartemisinin derivatives; malnutrition; received recentanti-malarial therapy that may affect the efficacy evalua-tion or any investigational drug within 30 days or fivehalf-lives (whichever longer); previously participated inthe study. A negative pregnancy test was required fromwomen of child-bearing age; breastfeeding mothers wereexcluded.

ProceduresDuring screening, all patients provided a full medical his-tory and underwent a comprehensive clinical examina-tion. Drug treatments were CDA 2/2.5/4 mg/kg, CPG-DDS 2/2.5 mg/kg (both GlaxoSmithKline, Greenford,UK) given once daily for three days (Days 0, 1, 2) or six-dose AL (Novartis Pharma AG, Basel, Switzerland). Ran-domization was 2:1 CDA versus the comparator. Detailsof patient assessments are given in the clinical reports ofthese trials [10,11]. Patients remained hospitalizedbetween Days 0 to 3, with follow up on Days 7, 14, and 28in both studies, and follow up also at Day 42 in the CDAversus AL study (005); with a home visit from a fieldwor-ker on Days 4, 5, and 6.

BlindingG6PD genotype and phenotype were determined by labora-tory staff blinded to treatment. Clinical investigators wereblinded to drug treatment and G6PD status was not avail-able to clinical investigators until after the trial wascomplete.

Parasite assessment>Asexual parasite and gametocyte counts were deter-mined by examination of duplicate Giemsa-stained thickblood slides (10 μL thumb prick) using WHO (2003)methods [14]. Thick blood slides were examined by twomicroscopists blinded to treatment; a third microscopistread the slides in case of a discrepancy. To differentiaterecrudescence from reinfection, two drops of blood col-lected at enrollment and subsequent date of parasitedetection were stored on filter paper for polymerasechain reaction (PCR) analysis of parasite genotype usingmsp-1, msp-2 and glurp [15,16].

G6PD genotypingG6PD genotype analysis was performed at two labora-tories: Walter Reed Project/Kenya Medical Research Insti-tute, Kisumu, Kenya and Shoklo Malaria Research Unit,Mae Sot, Thailand. After patient randomization, butbefore the first dose of study therapy, a 10 μL blood sam-ple was collected (thumb prick) onto pre-printed filterpaper. The G6PD genotyping protocol was based on pub-lished methods [17-19]. Specific forward primers andreverse primers were used for PCR amplification of a sec-tion of G6PD gene containing G6PD*B (wild type), thecommon African mutation G6PD*A (A376G) and theother deficient variant G6PD*A- (G202A, A542T, G680Tand T968C). Amplicons were subsequently analysed byrestriction fragment length polymorphism (RFLP) andgenotypes scored after agarose gel electrophoresis. DNAwas extracted using standard methods. PCR was per-formed using 0.75 U/μL Taq Polymerase, 1.5 mM MgCl2,200 μM dNTP (mix 10 mM), 2.5 μM of each primer, 10%PCR buffer, and 2.0 μL of DNA. All restriction endonu-cleases were sourced from New England Biolabs (Ipswich,MA, USA) and used at optimal temperature and digestionperiod. Genomic DNA was first amplified using primersfor A376G, 5’-CCCAGGCCACCCCAGAGGAGA-3’ (for-ward) and 5’-CGGCCCCGGACACGCTCATAG-3’(reverse). Thermocycling was performed at 94°C for 10min, then 30 cycles at 94°C for 45 sec, 58°C (annealingtemperature) for 45 sec, and 72°C for 45 sec, with a finalextension at 72°C for 7 min. Amplification products wererecovered and PCR fragments incubated at 37°C for 16 hwith FokI for identification of the A376G mutation(G6PD*A). All samples positive for A376G were then sub-jected to PCR amplification using: G202A, 5’-CCAC-CACTGCCCCTGTGACCT-3’ (forward) and 5’-GGCCCTGACACCACCCACCTT-3’ (reverse) with annealing at65°C; G542T, 5’-AGGAGATGTGGTTGGACATCCGG-3’(forward) and 5’-ACTCCCCGAAGAGGGGT-3’ (reverse)with annealing at 67°C; G680T, 5’-ACATGTGGCCCCTGCACCAC-3’ (forward) and 5’-GTGACTGGCTCTGCCACCCTG-3’ (reverse) with annealing at 69°C; andT968C 5’-TCCCTGCACCCCAACTCAAC-3’ (forward)and 5’-CCAGTTCTGCCTTGCTGGGC-3’ (reverse) withannealing at 65°C. Amplicons were then incubated at 37°Cfor 16 h with NlaIII for identification of the G202A muta-tion, AccIII for G542T, BstNI for G680T, and NciI for theT968C mutation. Quality control was maintained by com-paring results from the two testing laboratories withdiscrepant results reanalysed.

G6PD phenotypingPre-dose venous blood samples (2 mL) were collectedfor G6PD phenotyping, refrigerated to 4°C and trans-ported within seven days to a central laboratory (Synexa,


Page 3 of 14

Cape Town, South Africa) for analysis using a commer-cial NADPH fluorescence test (Trinity Biotech, Wick-low, Ireland). This was a qualitative test designed todistinguish grossly G6PD-deficient samples from thosewith intermediate/normal enzyme activity. The reactionmixture containing glucose-6-phosphate+NADP (notfluorescent) and spots were made on the filter paper atthe beginning (zero-time) and 5, 10, and 20 min afterblood incubation with reagent mixture. The spots werevisually inspected under long-wavelength (320-420 nm)ultraviolet light. The observed rate of appearance ofbright fluorescence is proportional to the blood G6PDHactivity; normal samples fluoresce brightly, whereas defi-cient samples show little or no fluorescence [20].

Efficacy outcomesDay-28 parasite recrudescence and reinfection rates, differ-entiated using PCR genotyping as described above, wereused to evaluate whether G6PD genotype or phenotypehad an effect on anti-malarial efficacy.

Statistical methodsDescriptive statistics were used to report demographic andclinical characteristics for patients with G6PD genotype.G6PD genotype and phenotype was presented by countryand centre.All statistical analyses were performed post-hoc. There

was no formal sample size calculation for the analysesreported in this paper. Sample size was calculated for clini-cal endpoints as reported previously [10,11]. Analysis ofconformity with the Hardy-Weinberg equilibrium forfemale gene frequencies was analysed by centre, countryand overall. Expected female gene frequencies were calcu-lated based on male genotype prevalence (A-, A and B)and compared with observed values using Chi-squared.A logistic model was developed to investigate the

effect of G6PD genotype on the following baseline vari-ables: haemoglobin concentration, temperature, andasexual parasite count.The effect of G6PD genotype was evaluated using three

categories: normal; heterozygous; or deficient (hemizygousmales plus homozygous females). The effect of G6PD gen-otype on baseline haemoglobin concentration was mod-elled with terms for age category (< 5, ≥ 5- < 15, ≥ 15years), weight, centre, sex, temperature and parasitaemiacategory (four quartiles). The model for the effect of G6PDgenotype on temperature had terms for age category,weight, centre, sex, haemoglobin concentration and parasi-taemia category. To investigate the effect of G6PD geno-type on parasitaemia, parasite densities were initially logtransformed with model terms for age category, weight,centre, sex, haemoglobin concentration and temperature.A logistic model was developed to investigate the effect

of G6PD genotype (normal, heterozygous, deficient) on

anti-malarial efficacy (recrudescence and reinfectionrates) with terms for: anti-malarial treatment, study(CDA vs. AL; CDA vs. CPG-DDS), age category, weight,centre, sex, and parasitaemia category. The same modelwas used to investigate the effect of G6PD phenotype(normal, intermediate, deficient) on anti-malarial efficacy.The models were used to calculate adjusted mean values

for the variable under investigation and a p value calcu-lated to test any significant effect of model terms usinganalysis of variance. G6PD genotype was further analysedto compare G6PD normal with heterozygous and G6PDnormal with G6PD deficient; 95% confidence intervalswere calculated and p values generated using analysis ofvariance.

ResultsPatientsA summary of the patient population that was screened,randomized and analysed for G6PD status is shown inFigure 1. Across the two studies 2264 malaria patientswere enrolled. G6PD genotype and phenotype was avail-able for 2045 and 1319 subjects, respectively, and thiscomprised the primary analysis population.

G6PD deficiency prevalenceTable 1 shows G6PD gene frequencies for males (N =1018) and females (N = 1027) by country and trial centre.There were no G6PD data from Lagos, Nigeria. Of the2045 patients, 77.7% (1588/2045) were G6PD normal,13.3% (273/2045) were female heterozygous and 9.0%(184/2045) were G6PD deficient, i.e. 7.2% (147/2045)male hemizygous plus 1.8% (37/2045) female homozy-gous. All deficient G6PD*A- genotypes were A376G/G202A; there were no instances of A542T, G680T orT968C. Hardy-Weinberg equilibrium was demonstratedfor four of 14 centres: Ouagadougou (Burkina Faso), Kilifi(Kenya), Banambani (Mali) and Kiwangwa (Tanzania)(Table 1).G6PD phenotype results are shown in Table 2. Pheno-

type was available for 64.5% (1319/2045) of patients.Overall, 10.2% (134/1319) of patients were phenotypicallyG6PD deficient, 9.6% (127/1319) intermediate, and 80.2%(1058/1319) normal (Table 2). Of the G6PD*A- hemizy-gous males that had phenotype data available, 70/99(70.7%) were phenotypically G6PD deficient and 29/99(29.3%) had intermediate/normal phenotype (Table 3).Of the G6PD*A- homozygous females, 12/25 (48.0%) haddeficient phenotype and 13/25 (52.0%) had normal/inter-mediate phenotype (Table 3). Likewise, 39/1021 (3.1%)patients of normal G6PD genotype (males A or B,females AA, AB or BB) were classified as G6PD deficientand 68/1021 (6.7%) as intermediate, illustrating the fluid-ity of G6PD measurements. As expected, female hetero-zygous displayed a wide range of G6PD activity.


Page 4 of 14

G6PD genotype effect on baseline variablesBaseline demographic and clinical characteristics of thestudy participants are summarized in Table 4. Figure 2shows baseline haemoglobin categorized by G6PD geno-type and age. Logistic regression found no significanteffect of G6PD genotype on adjusted mean baseline hae-moglobin (p = 0.154). Comparison between G6PD

normal and heterozygous (p = 0.693), or between G6PDnormal and deficient genotype (p = 0.0559) showed nodifference in adjusted mean baseline haemoglobin. Notethat this population was selected to have a baseline hae-moglobin ≥ 70 g/L and this may have biased the results.Figures 3 and 4 show baseline temperature and parasi-

taemia categorized by G6PD genotype, respectively.

17,534 subjects screened:758 (4.3) Burkina Faso882 (5.0) Ghana5425 (30.9) Kenya318 (1.8) Mali7195 (41.0) Nigeria2956 (16.9) Tanzania

2264 randomised:365 (16.1) Burkina Faso256 (11.3) Ghana300 (13.3) Kenya310 (13.7) Mali701 (31.0) Nigeria332 (14.7) Tanzania

Genotype population2045 (90.3) with G6PD genotype:357 (17.5) Burkina Faso236 (11.5) Ghana293 (14.3) Kenya308 (15.1) Mali547 (26.7) Nigeria304 (14.9) Tanzania

Phenotype population1319 (58.3) with G6PD genotype andphenotype:343 (26.0) Burkina Faso138 (10.5) Ghana219 (16.6) Kenya294 (22.3) Mali164 (12.4) Nigeria161 (12.2) Tanzania

1372 randomisedto 005 study

914 randomised toCDA

1201 with G6PDgenotype in 005 study

844 with G6PDgenotype in 006 study

590 with G6PDphenotype in 005 study

729 with G6PDphenotype in 006 study

458 randomised toAL

292 randomised toCPG-DDS

600 randomised toCDA

800 patients inCDA group

401 patients inAL group

282 patients inCPG-DDS group



198 patients in ALgroup

All values are shown as n (%).

243 patients inCPG-DDS group


892 randomised to006 study

Figure 1 Trial profile and patient population.


Page 5 of 14

Table 1 Frequency of G6PD genotypes in malaria patients from 17 centres in six African countries

Country Centre Male N A- A B Female N A-A- AA- BA- AA BB BA HW* p

BF Bobo-Dioulasso 1 0 1 (100) 0 4 0 0 2 (50.0) 0 1 (25.0) 1 (25.0) NC

Ouagadougou 174 27 (15.5) 48 (27.6) 99 (56.9) 178 6 (3.4) 15 (8.4) 33 (18.5) 18 (10.1) 54 (30.3) 52 (29.2) 0.742

Total 175 27 (15.4) 49 (28.0) 99 (56.6) 182 6 (3.3) 15 (8.2) 35 (19.2) 18 (9.9) 55 (30.2) 53 (29.1) 0.771

Ghana Kintampo 92 16 (17.4) 52 (56.5) 24 (26.1) 87 8 (9.2) 10 (11.5) 21 (24.1) 9 (10.3) 22 (25.3) 17 (19.5) 0.003

Kumasi 19 2 (10.5) 4 (21.1) 13 (68.4) 38 1 (2.6) 1 (2.6) 5 (13.2) 6 (15.8) 13 (34.2) 12 (31.6) 0.019

Total 111 18 (16.2) 28 (25.2) 65 (58.6) 125 9 (7.2) 11 (8.8) 26 (20.8) 15 (12.0) 35 (28.0) 29 (23.2) 0.001

Kenya Eldoret 77 15 (19.5) 22 (28.6) 40 (51.9) 72 2 (2.8) 2 (2.8) 10 (13.9) 1 (1.4) 42 (58.3) 15 (20.8) < 0.001

Kilifi 41 8 (19.5) 10 (24.4) 23 (56.1) 33 2 (6.1) 3 (9.1) 8 (24.2) 1 (3.0) 10 (30.3) 9 (27.3) 0.961

Pingilikani 39 5 (12.8) 7 (17.9) 27 (69.2) 31 1 (3.2) 7 (22.6) 5 (16.1) 1 (3.2) 11 (35.5) 6 (19.4) < 0.001

Total 157 28 (17.8) 39 (24.8) 90 (57.3) 136 5 (3.7) 12 (8.8) 23 (16.9) 3 (2.2) 63 (46.3) 30 (22.1) 0.016

Mali Doneguebougou 114 10 (8.8) 41 (36.0) 63 (55.3) 102 2 (2.0) 4 (3.9) 20 (19.6) 14 (13.7) 39 (38.2) 23 (22.5) < 0.001

Banambani 48 7 (14.6) 13 (27.1) 28 (58.3) 44 1 (2.3) 1 (2.3) 7 (15.9) 2 (4.5) 23 (52.3) 10 (22.7) 0.176

Total 162 17 (10.5) 54 (33.3) 91 (56.2) 146 3 (2.1) 5 (3.4) 27 (18.5) 16 (11.0) 62 (42.5) 33 (22.6) < 0.001

Nigeria Calabar 76 11 (14.5) 11 (14.5) 54 (71.1) 69 0 7 (10.1) 13 (18.8) 5 (7.2) 33 (47.8) 11 (15.9) 0.005

Enugu 121 17 (14.1) 32 (26.4) 72 (59.5) 98 5 (5.1) 11 (11.2) 23 (23.5) 1 (1.0) 41 (41.8) 17 (17.3) 0.001

Ibadan 26 2 (7.7) 2 (7.7) 22 (84.6) 30 0 3 (10.0) 4 (13.3) 3 (10.0) 15 (50.0) 5 (16.7) < 0.001

Ile Ife 49 7 (14.3) 14 (28.6) 28 (57.1) 57 5 (8.8) 4 (7.0) 14 (24.6) 4 (7.0) 20 (35.1) 10 (17.5) 0.002

Jos 11 0 4 (36.4) 7 (63.6) 10 0 2 (20.0) 1 (10.0) 3 (30.0) 2 (20.0) 2 (20.0) NC

Total 283 37 (13.1) 63 (22.3) 183 (64.7) 264 10 (3.8) 27 (10.2) 55 (20.8) 16 (6.1) 111 (42.0) 45 (17.0) < 0.001

Tanzania Bagamoyo 19 4 (21.1) 4 (21.1) 11 (57.9) 23 0 0 0 5 (21.7) 13 (56.5) 5 (21.7) < 0.001

Kiwangwa 111 16 (14.4) 18 (16.2) 77 (69.4) 151 4 (2.6) 11 (7.3) 26 (17.2) 3 (2.0) 60 (39.7) 47 (31.1) 0.063

Total 130 20 (15.4) 22 (16.9) 88 (67.7) 174 4 (2.3) 11 (6.3) 26 (14.9) 8 (4.6) 73 (42.0) 52 (29.9) 0.094

Total 1018 147 (14.4) 283 (27.8) 588 (57.8) 1027 37 (3.6) 81 (7.9) 192 (18.7) 76 (7.4) 399 (38.8) 242 (23.5) < 0.001

Data shown as n (%) BF, Burkina Faso; NC, not calculated. G6PD genotype: male hemizygous = A-; male normal = A or B; female homozygous = A-/A-; female heterozygous = A/A- or B/A-; and female normal = A/A,B/B or B/A.

*p values < 0.05 indicate significant differences from predicted G6PD gene frequencies in females based on male gene frequencies for A-, A and B alleles using the Hardy-Weinberg equation (Chi-squared).

Carter

etal.M

alariaJournal2011,10:241

http://www.m

alariajournal.com/content/10/1/241

Page6of

14

Logistic regression showed no significant effect of G6PDgenotype on adjusted mean baseline temperature (p =0.9617). Comparisons of adjusted mean baseline tem-perature between G6PD normal and heterozygous (p =

0.784) or G6PD normal and deficient genotype (p =0.969) found no significant differences. G6PD genotypehad no significant effect on adjusted log mean baselineparasitaemia (p = 0.365). There was no difference inadjusted log mean parasitaemia between G6PD normalversus heterozygous (p = 0.181) or G6PD normal versusdeficient genotype (p = 0.693).

G6PD genotype and anti-malarial efficacyTable 5 shows recrudescence and reinfection rates bygenotype and phenotype. There were 104/2045 (5.1%)patients with parasite recrudescence during the CDAclinical trials. Logistic modelling found no effect ofG6PD genotype (p = 0.490) or phenotype (p = 0.391) onrecrudescence rate. There was also no significant differ-ence in recrudescence rate between normal and hetero-zygous G6PD genotypes (odds ratio [OR] 0.7; 95%CI0.4, 1.4; p = 0.315) or normal and deficient genotypes(OR 1.2; 95%CI 0.6, 2.4; p = 0.561).Reinfection with falciparum malaria occurred in 608/

2045 (29.7%) patients overall in the study. Based on theraw data (Table 5), there was a trend for lower reinfec-tion rates in patients that had G6PD-deficient (26.6%) orheterozygous genotypes (26.0%) versus normal genotype(30.7%). Also, patients with a deficient phenotype had alower reinfection rate (26.7%) than those with a normal

Table 2 Frequency of G6PD phenotype by centre and country

Country Centre N* Deficient Intermediate Normal

BF Bobo-Dioulasso 0 0 0 0

Ouagadougou 343 29 (8.5) 37 (10.8) 277 (80.8)

Total 343 29 (8.5) 37 (10.8) 277 (80.8)

Ghana Kintampo 118 8 (6.8) 15 (12.7) 95 (80.5)

Kumasi 20 1 (5.0) 0 19 (95.0)

Total 138 9 (6.5) 15 (10.9) 114 (82.6)

Kenya Eldoret 90 15 (16.7) 4 (4.4)1 71 (78.9)

Kilifi 67 7 (10.4) 7 (10.4) 53 (79.1)

Pingilikani 62 4 (6.5) 3 (4.8) 55 (88.7)

Total 219 26 (11.9) 14 (6.4) 179 (81.7)

Mali Doneguebougou 205 17 (8.3) 20 (9.8) 168 (82.0)

Banambani 89 10 (11.2) 11 (12.4) 68 (76.4)

Total 294 27 (9.2) 31 (10.5) 236 (80.3)

Nigeria Calabar 21 5 (23.8) 2 (9.5) 14 (66.7)

Enugu 71 6 (8.5) 10 (14.1) 55 (77.5)

Ibadan 0 0 0 0

Ile Ife 72 7 (9.7) 4 (5.6) 61 (84.7)

Jos 0 0 0 0

Total 164 18 (11.0) 16 (9.8) 130 (79.3)

Tanzania Bagamoyo 10 2 (20.0) 0 8 (80.0)

Kiwangwa 151 23 (15.2) 14 (9.3) 114 (75.5)

Total 161 25 (15.5) 14 (8.7) 122 (75.8)

Total 1319 134 (10.2) 127 (9.6) 1058 (80.2)

Frequencies are shown as n (%).

*N is the number of patients with genotype and phenotype data.

Table 3 Relationship between phenotype and genotypetesting

G6PD genotype G6PD phenotype n/N (%)

Male A- Deficient 70/99 (70.7)

Intermediate 19/99 (19.2)

Normal 10/99 (10.1)

Male A or B Deficient 25/575 (4.3)


Normal 522/575 (90.8)

Female A-/A- Deficient 12/25 (48.0)


Normal 6/25 (24.0)

Female A-/A or A-/B** Deficient 13/174 (7.5)


Normal 128/174 (73.6)

Female AA, AB or BB Deficient 14/446 (3.1)


Normal 392/446 (87.9)

G6PD genotype: male hemizygous = A-; male normal = A or B; femalehomozygous = A-/A-; female heterozygous = A/A- or B/A-; and female normal= A/A, B/B or B/A.


Page 7 of 14

phenotype (35.7%). However, within the logistic modelthese relationships failed to reach statistical significanceoverall: for genotype p = 0.134 and for phenotype p =0.354. The difference in reinfection rate also failed toreach statistical significance between normal and hetero-zygous G6PD genotypes (OR 0.8; 95%CI 0.5, 1.1; p =0.105) or normal and deficient genotypes (OR 0.8; 95%CI0.5, 1.1; p = 0.203).

DiscussionThe primary objective of this study was to describe theprevalence of G6PD genotypes in African malariapatients who participated in two CDA Phase III clinicaltrials. In this study, G6PD*A- prevalence varied betweentrial centres and between countries (Table 1). Wherecomparable data exists for each country, the prevalenceof G6PD deficiency in the current study was lower com-pared with previous reports from Burkina Faso (31.0%)[21], and Nigeria (21.6%, 23.9% and 24.2%) [22-24], buthigher than reported for Ghana (8.5%) [25]. A study inMali of children with uncomplicated or complicated

malaria, reported a G6PD*A- prevalence of 12.9% [5];about the same as reported here (10.5%). There are noprevious reports for G6PD genotype prevalence usingDNA analysis for Kenya or Tanzania for comparison.The current study selected for particular malaria

patients, so it is not surprising that G6PD gene frequen-cies are divergent from those reported for randomlyselected healthy individuals in the general population.Notably, patients with known G6PD deficiency or neo-natal hyperbilirubinaemia were excluded. Unfortunately,data are not available on how many patients wereexcluded from the clinical trials because of knownG6PD deficiency. However, anecdotally, there wereprobably very few as the regions studied were naïve toG6PD testing. Conversely, G6PD prevalence data from ageneral population cannot necessarily be used to esti-mate the proportion of G6PD-deficient individuals thatwill be enrolled into an anti-malarial clinical trial. Fromthe postulated evolution of G6PD deficiency, it is possi-ble that the frequency of malaria is lower in patientswith G6PD deficiency, though a protective effect against

Table 4 Baseline demographic and clinical data by G6PD genotype

Characteristic* Malehemizygous(n = 147)

Malenormal(n = 871)

Femalehomozygous

(n = 37)

Femaleheterozygous(n = 273)

Femalenormal(n = 717)

All patients(n = 2045)

Age, years [range] 5.8 (8.0) [1-70] 5.1 (5.5) [1-59]

7.9 (10.0) [1-42] 5.6 (7.8) [1-72] 5.6 (6.5) [1-72]

5.44 (6.5) [1-72]

1- < 5, n (%) 87 (59) 511 (59) 21 (57) 169 (62) 412 (57) 1200 (59)

5- < 15, n (%) 52 (35) 331 (38) 10 (27) 90 (33) 277 (39) 760 (37)

≥ 15, n (%) 8 (5) 29 (3) 6 (16) 14 (5) 28 (4) 85 (4)

Weight, kg, mean 18.8 (11.7) 17.9 (10.5) 23.1 (21.9) 18.0 (12.2) 18.3 (11.0) 18.2 (11.3)

Parasitaemia, μL-1, geometric mean[range]

25990[1484-705600]

26331[216-

323361]

18602[1026-211546]

22498[0-303400]

27213[0-389415]

25895[0-705600]

Temperature, °C 37.8 (1.0) 37.9 (1.0) 37.9 (1.0) 37.8 (1.0) 37.9 (1.0) 37.9 (1.0)

Fever, n (%) 81 (55.1) 552 (63.4) 24 (65.0) 165 (60.0) 459 (64.0) 1281 (62.6)

Haemoglobin, g/L* 100.3 (18.3) 101.1 (16.8) 99.7 (19.2) 100.6 (14.2) 102.7 (15.6) 101.5 (16.2)

1 to < 5 years 93.2 (15.2) 95.6 (14.6) 91.0 (16.5) 96.1 (12.6) 97.3 (14.6) 96.0 (14.5)

5 to < 15 years 107.2 (15.2) 106.6 (14.8) 104.4 (12.5) 105.8 (13.2) 108.9 (13.9) 107.3 (14.3)

≥ 15 years 132.5 (16.5) 135.7 (12.2) 122.5 (17.5) 119.9 (11.9) 120.5 (9.8) 126.9 (14.0)

Haematocrit, L 0.30 (0.05) 0.30 (0.05) 0.31 (0.05) 0.31 (0.04) 0.31 (0.04) 0.31 (0.05)

RBC count x1012/L 4.0 (0.7) 4.2 (0.6) 4.0 (0.6) 4.0 (0.5) 4.1 (0.6) 4.1 (0.62)

Platelet count x109/L 191.4 (86.5) 196.7(111.6)

206.1 (125.2) 186.9 (96.4) 198.1 (128.0) 195.7 (114.5)

WBC count x109/L 9.6 (4.3) 9.5 (4.1) 8.5 (4.2) 9.5 (3.9) 9.4 (4.0) 9.5 (4.0)

Treatment n (%)

CDA 005 64 (43.5) 324 (37.2) 16 (43.2) 123 (45.1) 273 (38.1) 800 (39.1)

CDA 006 38 (25.9) 250 (28.7) 6 (16.2) 70 (25.6) 198 (27.6) 562 (27.5)

AL 30 (20.4) 185 (21.2) 6 (16.2) 43 (15.8) 137 (19.1) 401 (19.6)

CPG-DDS 15 (10.2) 112 (12.9) 9 (24.3) 37 (13.6) 109 (15.2) 282 (13.8)

All values are mean (SD) [range] unless otherwise indicated. CDA, chlorproguanil-dapsone-artesunate; AL, artemether-lumefantrine; CPG-DDS, chlorproguanil-dapsone; RBC, red blood cell; WBC, white blood cell. G6PD genotype: male hemizygous = A-; male normal = A or B; female homozygous = A-/A-; femaleheterozygous = A/A- or B/A-; and female normal = A/A, B/B or B/A.

*There were 23 patients with baseline haemoglobin of < 70 g/L (range 45-69 g/L, mean 65 g/L). This occurred because at some sites screening haemoglobinlevel was determined using a rapid test (Haemacue, Downfield, UK), with subsequent checking using a coulter counter the results of which are reported here.


Page 8 of 14

40

50

60

70

80

90

100

110

120

130

140

150

160

170H

aem

og

lob

in (

g/L

)

40

50

60

70

80

90

100

110

120

130

140

150

160

170

Hae

mo

glo

bin

(g

/L)

40

50

60

70

80

90

100

110

120

130

140

150

160

170

Hae

mo

glo

bin

(g

/L)

Lower quartile Upper quartile Minimum MedianMaximum

Male hemizygous(n=8)

Male normal(n=29)

Female homozygous(n=6)

G6PD status

Female heterozygous(n=14)

Female normal(n=28)

A

B

C

(n=87) (n=511) (n=21) (n=169) (n=412)

(n=52) (n=331) (n=10) (n=90) (n=277)

Figure 2 Baseline haemoglobin by G6PD genotype for patients aged: A, 1- < 5 years; B, 5- < 15 years; and C, ≥ 15 years. G6PDgenotype: male hemizygous = A-; male normal = A or B; female homozygous = A-/A-; female heterozygous = A/A- or B/A-; and female normal =A/A, B/B or B/A.


Page 9 of 14

uncomplicated malaria has not been conclusivelydemonstrated.In this study, all the G6PD*A- mutations involved

A376G/G202A, confirming this variant as the most com-mon in Africa. The A542T, G680T or T968C G6PD*A-mutations were not detected in the malaria patientsincluded in this study. Data on the frequency of thesemutations are sparse in Africa. The 968C mutation hasbeen reported as the most common G6PD*A- allele inThe Gambia [3] and in the Sereer ethnic group fromSenegal [8]. Given the selected population in this study itcannot be confirmed whether any of these mutations arenot present in the countries studied. The limitations ofthe assay regarding other possible G6PD-deficient muta-tions that were not tested for in this study should also beconsidered. Further genotyping studies are required todetermine the frequency of these mutations in the gen-eral population and specific ethnic groups.The Hardy-Weinberg equation describes a population in

which both allele and genotype frequencies do not change,meaning they are in equilibrium. Using this equation, tencentres had female gene frequencies significantly differentto those predicted from male gene frequencies (Table 1).The sampling methods in the studied population (as deter-mined by the study inclusion/exclusion criteria) may not

have enrolled the different genotypes at the same frequen-cies that they occur in the general population. Also, thesample size was not large enough in some centres to con-firm or reject the hypothesis of genetic equilibrium. Otherexplanations include changes in selection pressure, such asfrom malaria control or anti-malarial treatment, or popula-tion mixing. Evidence from Taiwan indicates that recentimmigration can cause changes in relative G6PD gene fre-quencies between males and females independent of theoverall incidence of G6PD deficiency [26]. This may berelevant for some of the centres in this study; for example,Eldoret is one of the fastest growing towns in Kenya, andwith 250 different ethnic groups in the country, it wouldbe unlikely if genetic equilibrium for G6PD genotypes wasmaintained. However, it is difficult to separate the complexreasons for the divergence from Hardy-Weinberg equili-brium and further specific investigations would berequired. There is also a methodological explanation. Inmany studies on G6PD-deficiency variant prevalence, onlythe frequency of the A- allele is known and this alone isused to derive predicted female genotype frequencies forG6PD deficient, heterozygous and normal; thus, only theA- allele is tested for equilibrium. In this study, the fre-quencies of the A-, A and B alleles in males were used tocalculate predicted frequencies of the female genotypes,

34

35

36

37

38

39

40

41

Tem

per

atu

re (

ºC)


Male normal(n=871)


G6PD status


Female normal(n=717)

Lower quartile

Upper quartile

Minimum Median

Maximum

Figure 3 Baseline temperature by G6PD genotype. G6PD genotype: male hemizygous = A-; male normal = A or B; female homozygous = A-/A-; female heterozygous = A/A- or B/A-; and female normal = A/A, B/B or B/A.


Page 10 of 14

A-A-, AA-, BA-, AA, BB and BA, which is a strict applica-tion of the Hardy-Weinberg equation and so less likely toshow conformance unless all G6PD alleles and genotypesare in equilibrium.G6PD genotyping has limitations, for instance, it is not

practical as a routine test in a clinical setting. Also,

because the gene is highly polymorphic (more than 400reported variants), unusual but clinically important var-iants can be missed [2]. Phenotyping using the NADPHfluorescence spot test is an alternative test as it detectsdeficiency irrespective of underlying gene mutations(Table 2). The biggest limitation of G6PD phenotyping isits inability to conclusively identify heterozygous femalesbecause their G6PD levels can range from near normalto deficient. As shown in Table 3, 73.6% (128/174) offemale heterozygous patients were classified as ‘normal’ byG6PD phenotyping. In the CDA trials, heterozygousfemales appeared to have no greater risk for clinicallysignificant haemolysis than normal females (haemoglobindecrease of ≥ 40 g/L or ≥ 40% versus baseline or haemo-globin < 50 g/L or blood transfusion) [10,11]. However,this cannot be assumed for all drugs.As shown in Table 3, the specificity of phenotype test-

ing for identifying G6PD*A- was 66.1% (82/124), i.e.33.9% (42/124) of G6PD*A- patients were not identifiedby phenotype. This misclassification is probably attribu-table to reticulocyte rebound subsequent to haemolyticcrisis, or recovery from malaria [9]. In patients withmalaria, higher than expected G6PD enzyme levels canoccur in individuals with G6PD-deficient genotypesbecause the increased erythrocyte replacement rate

Table 5 Malaria recrudescence and reinfection rates atDay 28 by G6PD genotype and phenotype

G6PD status N Recrudescencen (%)

Reinfectionn (%)

Genotype*

Deficient 184 11 (6.0) 49 (26.6)

Heterozygous 273 11 (4.0) 71 (26.0)

Normal 1588 82 (4.0) 488 (30.7)

Total 2045 104 (5.1) 608 (29.7)

Phenotype**

Deficient 146 9 (6.2) 39 (26.7)

Intermediate 137 3 (2.2) 39 (28.5)

Normal 1096 62 (5.7) 391 (35.7)

Total 1379 74 (5.4) 469 (34.0)

*G6PD genotype: male hemizygous = A-; male normal = A or B; femalehomozygous = A-/A-; female heterozygous = A/A- or B/A-; and female normal =A/A, B/B or B/A.

**Analysis performed for patients with G6PD phenotype with or withoutgenotype available.

2.00

4.00

5.00

3.00

6.00

Ase

xual

par

asit

e co

un

t (μ

L-1)


Male normal(n=871)


G6PD status


Female normal(n=717)

Lower quartile

Upper quartile

Minimum Median

Maximum

Figure 4 Baseline parasitaemia (Log10) by G6PD genotype. G6PD genotype: male hemizygous = A-; male normal = A or B; femalehomozygous = A-/A-; female heterozygous = A/A- or B/A-; and female normal = A/A, B/B or B/A.


Page 11 of 14

results in a younger erythrocyte population and newlyformed erythrocytes have a greater capacity for G6PDproduction [9]. However, for unknown reasons, the cor-relation between deficient phenotype and G6PD*A- gen-otype was better for hemizygous males, specificity 70.7%(70/99), than for homozygous females, specificity 48.0%(12/25). This has also been seen in a recent paper where83.3% (30/36) of hemizygous males were phenotypicallydeficient whereas specificity was only 60.0% (3/5) inhomozygous females, though numbers in the study werevery low compared with the current dataset [27]. Inmales and females with normal G6PD genotype (A, B,AA or AB), 4.3% (25/575) and 3.1% (14/446), respec-tively, were found to be phenotypically deficient. It ispossible that these individuals might have had G6PDmutations at loci other than those examined.In the context of this trial, using only phenotype data

would have significantly undermined interpretation of thedrug-induced G6PD-related haemolytic effect of CDA asabout a third of G6PD*A- patients would be expected tohave had a clinically important haemolysis [10,11].Because of this, and the limitation in identifying heterozy-gous females, G6PD genotyping will continue to be impor-tant, particularly in a research situation. In the widercontext, it has been suggested that phenotype could beused to exclude G6PD*A- patients from receiving poten-tially harmful treatments. It is possible to calculate thenumber of G6PD*A- patients that would have been inad-vertently treated after phenotype screening (Table 3).Excluding the 134 patients who had a deficient phenotypewould leave a total population of 1185 exposed to treat-ment, 42 (3.5%) of whom were of G6PD*A- genotype.Excluding patients with a deficient or intermediate pheno-type (n = 261), leaves a total population of 1058, of whom16 (1.5%) had a G6PD*A- genotype. The number ofG6PD*A- patients misclassified by phenotype appearssmall (1.5%), i.e. about five patients would have had signifi-cant hemolysis that would have been avoidable with geno-typing. However, where alternative therapies exist, in theclinical setting it would be unethical to expose even asmall proportion of patients with a G6PD-deficient geno-type to potentially life-threatening haemolysis, particularlywhere patient follow up is limited.Across all patients, phenotype correlated reasonably well

with the occurrence of significant hemolysis. Of the 878patients with phenotype data, significant hemolysisoccurred in 21.1% (19/90) of those patients who were phe-notypically deficient, in 5.6% (5/90) of those who wereintermediate and in 1.1% (8/698) of those with normalphenotype. Of the 41/124 (33.0%) G6PD*A- patients trea-ted with CDA who had clinically important haemolysis, 23had phenotype data available. Of these, 17/23 (73.9%)were phenotypically deficient, 4/23 (17.4%) had intermedi-ate phenotype and 2/23 (8.7%) had normal phenotype.

The study design excluded patients with haemoglobinconcentrations ≤ 70 g/L and this consideration may haveaffected the study results. There were no independenteffects of G6PD genotype on baseline haemoglobin, tem-perature, or malaria parasitaemia in the study population(Table 4, Figures 2, 3 and 4). Thus, judged by these para-meters, G6PD-deficient malaria patients appeared to be nodifferent from other malaria patients at presentation.These results are consistent with a study in Tanzania inmalaria patients that showed no effect of G6PD genotypeon haemoglobin concentration or parasite density [28].Also, in Nigerian subjects with no or asymptomaticmalaria, G6PD deficiency had no significant effect on hae-moglobin levels, though red blood cell concentrationswere higher in G6PD-normal subjects versus those whowere G6PD deficient [23].As shown in Table 5, there was a trend for patients with

G6PD heterozygous or deficient genotypes or G6PD defi-cient phenotype to have lower unadjusted reinfectionrates. However, after adjustment for baseline factors,including drug treatment, no effect of G6PD genotype orphenotype on recrudescence or reinfection rates wasfound. A recent study found no effect of G6PD deficiencyon P. falciparum clearance after treatment with artemisi-nin-based combination therapy [29]. However, there issome evidence that G6PD deficiency may reduce the num-ber of parasite strains carried by individuals in transmis-sion zones/seasons [30].There are limitations in performing post-hoc statistical

analysis on the effect of G6PD status on baseline vari-ables and efficacy as was done in this study. Logisticmodelling also has limitations; inclusion of the variablesfor adjustment is somewhat subjective and can never becomprehensive. Therefore, the conclusions that can bedrawn from this study should be viewed as exploratoryand require verification in prospective studies.

ConclusionsThe G6PD*A- deficient genotype was relatively commonamong the diverse African populations included in thisstudy. This represents a significant risk for adverse hae-molytic events after treatment with drug therapies thathave the potential to induce oxidative stress. G6PD geno-typing and phenotyping should therefore be a requisite inclinical trials evaluating the safety and efficacy of suchdrugs. The G6PD prevalence data presented here arerelevant for the design of such anti-malarial clinical trials.Lastly, there were no clinical differences between G6PD-deficient and -normal patients with malaria enrolled intothe CDA clinical trials.

AcknowledgementsNaomi Richardson of Magenta Communications Ltd developed the first draftof this paper and collated author comments and was funded by


Page 12 of 14

GlaxoSmithKline PLC. We are thankful to Lab 1 team at Walter Reed Project/KEMRI, Kisumu, Kenya for performing the G6PD genotype assays. Theauthors thank Dr François Nosten and his team at the Shoklo MalariaResearch Unit, Mae Sot, Thailand for performing quality control for the G6PDanalysis. Thanks also to the Medicines for Malaria Venture for significantfunding of the phenotype and genotype analysis.Funding statementThe CDA clinical trial programme was conducted through a public-privatepartnership within a development agreement between the Medicines forMalaria Venture, the World Health Organization Special Programme forResearch and Training in Tropical Diseases and GlaxoSmithKline PLC.Liverpool School of Tropical Medicine, Liverpool University, and the LondonSchool of Hygiene and Tropical Medicine joined the development team asacademic partners. A committee representing members of the CDA JointDevelopment Team and study site Principal Investigators developed theprotocol. GlaxoSmithKline conducted the study, collected and analysed data.All authors had access to the primary data and take responsibility for datareporting accuracy and completeness. The corresponding author hadresponsibility for the final decision to submit for publication.

Author details1ID-MDC Biomedical Data Sciences, GlaxoSmithKline Research andDevelopment, Stockley Park West, Uxbridge, Middlesex, UB11 1BT, UK.2Developing Countries & Market Access, GlaxoSmithKline, GSK House CN608, 980 Great West Road, Brentford, Middlesex, TW8 9GS, UK. 3Medicines forMalaria Venture, International Centre Cointrin, 20 Route de Pré-Bois, 1215Geneva 15, Switzerland. 4Walter Reed Project/Kenya Medical ResearchInstitute, Kisumu, United Nations Avenue Gigiri, Village Market, Nairobi 00621,Kenya.

Authors’ contributionsAll authors contributed to the study design. Genotype testing and scoringwas performed in JW’s laboratory. NC conducted the statistical analysis. Allauthors contributed to the preparation of the manuscript and approved thefinal version.

Competing interestsNick Carter and Allan Pamba are employees of GlaxoSmithKline. StephanDuparc is an employee of the Medicines for Malaria Venture.

Received: 19 May 2011 Accepted: 17 August 2011Published: 17 August 2011

References1. Luzzatto L, Mehta A, Vulliamy T: Glucose-6-phosphate dehydrogenase

deficiency. In The Metabolic & Molecular Bases of Inherited Disease. Volume 3.Edited by: Scriver C, Beaudet A, Sly W, Valle D. New York: McGraw Hill;2001:4517-4553.

2. Beutler E, Vulliamy TJ: Hematologically important mutations: glucose-6-phosphate dehydrogenase. Blood Cells Mol Dis 2002, 28:93-103.

3. Clark TG, Fry AE, Auburn S, Campino S, Diakite M, Green A, Richardson A,Teo YY, Small K, Wilson J, Jallow M, Sisay-Joof F, Pinder M, Sabeti P,Kwiatkowski DP, Rockett KA: Allelic heterogeneity of G6PD deficiency inWest Africa and severe malaria susceptibility. Eur J Hum Genet 2009,17:1080-1085.

4. Ruwende C, Khoo SC, Snow RW, Yates SN, Kwiatkowski D, Gupta S, Warn P,Allsopp CE, Gilbert SC, Peschu N, Newbold C, Greenwood B, Marsh K, Hill A:Natural selection of hemi- and heterozygotes for G6PD deficiency inAfrica by resistance to severe malaria. Nature 1995, 376:246-249.

5. Guindo A, Fairhurst RM, Doumbo OK, Wellems TE, Diallo DA: X-linkedG6PD deficiency protects hemizygous males but not heterozygousfemales against severe malaria. PLoS Med 2007, 4:e66.

6. Vulliamy TJ, Othman A, Town M, Nathwani A, Falusi AG, Mason PJ,Luzzatto L: Polymorphic sites in the African population detected bysequence analysis of the glucose-6-phosphate dehydrogenase geneoutline the evolution of the variants A and A-. Proc Natl Acad Sci USA1991, 88:8568-8571.

7. Hirono A, Beutler E: Molecular cloning and nucleotide sequence of cDNAfor human glucose-6-phosphate dehydrogenase variant A(-). Proc NatlAcad Sci USA 1988, 85:3951-3954.

8. De Araujo C, Migot-Nabias F, Guitard J, Pelleau S, Vulliamy T, Ducrocq R:The role of the G6PD AEth376G/968C allele in glucose-6-phosphatedehydrogenase deficiency in the seerer population of Senegal.Haematologica 2006, 91:262-263.

9. Beutler E, Duparc S: Glucose-6-phosphate dehydrogenase deficiency andantimalarial drug development. Am J Trop Med Hyg 2007, 77:779-789.

10. Tiono AB, Dicko A, Ndububa DA, Agbenyega T, Pitmang S, Awobusuyi J,Pamba A, Duparc S, Goh LE, Harrell E, Carter N, Ward SA, Greenwood B,Winstanley PA: Chlorproguanil-dapsone-artesunate versus chlorproguanil-dapsone: a randomized, double-blind, phase III trial in African children,adolescents, and adults with uncomplicated Plasmodium falciparummalaria. Am J Trop Med Hyg 2009, 81:969-978.

11. Premji Z, Umeh RE, Owusu-Agyei S, Esamai F, Ezedinachi EU, Oguche S,Borrmann S, Sowunmi A, Duparc S, Kirby PL, Pamba A, Kellam L,Guiguemde R, Greenwood B, Ward SA, Winstanley PA: Chlorproguanil-dapsone-artesunate versus artemether-lumefantrine: a randomized,double-blind phase III trial in African children and adolescents withuncomplicated Plasmodium falciparum malaria. PLoS One 2009, 4:e6682.

12. Luzzatto L: The rise and fall of the antimalarial Lapdap: a lesson inpharmacogenetics. Lancet 376:739-741.

13. Alloueche A, Bailey W, Barton S, Bwika J, Chimpeni P, Falade CO,Fehintola FA, Horton J, Jaffar S, Kanyok T, Kremsner PG, Kublin JG, Lang T,Missinou MA, Mkandala C, Oduola AM, Premji Z, Robertson L, Sowunmi A,Ward SA, Winstanley PA: Comparison of chlorproguanil-dapsone withsulfadoxine-pyrimethamine for the treatment of uncomplicatedfalciparum malaria in young African children: double-blind randomisedcontrolled trial. Lancet 2004, 363:1843-1848.

14. World Health Organization: Assessment and monitoring of antimalarialdrug efficacy for the treatment of uncomplicated falciparum malaria(WHO/HTM/RBM/2003.50). [http://www.who.int/malaria/publications/atoz/whohtmrbm200350/en/].

15. Cattamanchi A, Kyabayinze D, Hubbard A, Rosenthal PJ, Dorsey G:Distinguishing recrudescence from reinfection in a longitudinalantimalarial drug efficacy study: comparison of results based ongenotyping of msp-1, msp-2, and glurp. Am J Trop Med Hyg 2003,68:133-139.

16. Medicines for Malaria Venture, World Health Organization: Methods andtechniques for clinical trials on anti-malarial drug efficacy: genotyping toidentify parasite populations. [http://www.who.int/malaria/publications/atoz/9789241596305/en/].

17. Beutler E, Kuhl W, Vives-Corrons JL, Prchal JT: Molecular heterogeneity ofglucose-6-phosphate dehydrogenase A. Blood 1989, 74:2550-2555.

18. Nafa K, Reghis A, Osmani N, Baghli L, Ait-Abbes H, Benabadji M, Kaplan JC,Vulliamy T, Luzzatto L: At least five polymorphic mutants account for theprevalence of glucose-6-phosphate dehydrogenase deficiency in Algeria.Hum Genet 1994, 94:513-517.

19. Samilchuk E, D’Souza B, Al-Awadi S: Population study of commonglucose-6-phosphate dehydrogenase mutations in Kuwait. Hum Hered1999, 49:41-44.

20. Beutler E, Mitchell M: Special modifications of the fluorescent screeningmethod for glucose-6-phosphate dehydrogenase deficiency. Blood 1968,32:816-818.

21. Meissner PE, Coulibaly B, Mandi G, Mansmann U, Witte S, Schiek W,Muller O, Schirmer RH, Mockenhaupt FP, Bienzle U: Diagnosis of red cellG6PD deficiency in rural Burkina Faso: comparison of a rapid fluorescentenzyme test on filter paper with polymerase chain reaction basedgenotyping. Br J Haematol 2005, 131:395-399.

22. Luzzatto L, Allan NC: Relationship between the genes for glucose-6-phosphate dehydrogenase and for haemoglobin in a Nigerianpopulation. Nature 1968, 219:1041-1042.

23. May J, Meyer CG, Grossterlinden L, Ademowo OG, Mockenhaupt FP,Olumese PE, Falusi AG, Luzzatto L, Bienzle U: Red cell glucose-6-phosphatedehydrogenase status and pyruvate kinase activity in a Nigerianpopulation. Trop Med Int Health 2000, 5:119-123.

24. Ademowo OG, Falusi AG: Molecular epidemiology and activity oferythrocyte G6PD variants in a homogeneous Nigerian population. EastAfr Med J 2002, 79:42-44 [http://www.ajol.info/index.php/eamj/article/viewFile/8924/1751].

25. Burchard GD, Browne EN, Sievertsen J, May J, Meyer CG: Spleen sizedetermined by ultrasound in patients with sickle cell trait, HbAC trait


Page 13 of 14

http://www.ncbi.nlm.nih.gov/pubmed/21936056?dopt=Abstract


































http://www.who.int/malaria/publications/atoz/whohtmrbm200350/en/

http://www.who.int/malaria/publications/atoz/whohtmrbm200350/en/




http://www.who.int/malaria/publications/atoz/9789241596305/en/

http://www.who.int/malaria/publications/atoz/9789241596305/en/





















http://www.ajol.info/index.php/eamj/article/viewFile/8924/1751

http://www.ajol.info/index.php/eamj/article/viewFile/8924/1751



and glucose-6-phosphate-dehydrogenase deficiency in a malariahyperendemic area (Ashanti Region, Ghana). Acta Trop 2001, 80:103-109.

26. Chien YH, Lee NC, Wu ST, Liou JJ, Chen HC, Hwu WL: Changes inincidence and sex ratio of glucose-6-phosphate dehydrogenasedeficiency by population drift in Taiwan. Southeast Asian J Trop MedPublic Health 2008, 39:154-161 [http://www.tm.mahidol.ac.th/seameo/2008_39_1/21-4176.pdf].

27. Johnson MK, Clark TD, Njama-Meya D, Rosenthal PJ, Parikh S: Impact of themethod of G6PD deficiency assessment on genetic association studiesof malaria susceptibility. PLoS One 2009, 4:e7246.

28. Enevold A, Lusingu JP, Mmbando B, Alifrangis M, Lemnge MM, Bygbjerg IC,Theander TG, Vestergaard LS: Reduced risk of uncomplicated malariaepisodes in children with alpha+-thalassemia in northeastern Tanzania.Am J Trop Med Hyg 2008, 78:714-720.

29. Kone AK, Sagara I, Thera MA, Dicko A, Guindo A, Diakite S, Kurantsin-Mills J,Djimde A, Walcourt A, Doumbo O: Plasmodium falciparum clearance withartemisinin-based combination therapy (ACT) in patients with glucose-6-phosphate dehydrogenase deficiency in Mali. Malar J 9:332.

30. Vafa M, Troye-Blomberg M, Anchang J, Garcia A, Migot-Nabias F:Multiplicity of Plasmodium falciparum infection in asymptomatic childrenin Senegal: relation to transmission, age and erythrocyte variants. MalarJ 2008, 7:17.

doi:10.1186/1475-2875-10-241Cite this article as: Carter et al.: Frequency of glucose-6-phosphatedehydrogenase deficiency in malaria patients from six African countriesenrolled in two randomized anti-malarial clinical trials. Malaria Journal 201110:241.

Submit your next manuscript to BioMed Centraland take full advantage of:

• Convenient online submission

• Thorough peer review

• No space constraints or color figure charges

• Immediate publication on acceptance

• Inclusion in PubMed, CAS, Scopus and Google Scholar

• Research which is freely available for redistribution

Submit your manuscript at www.biomedcentral.com/submit


Page 14 of 14






http://www.tm.mahidol.ac.th/seameo/2008_39_1/21-4176.pdf

http://www.tm.mahidol.ac.th/seameo/2008_39_1/21-4176.pdf











Documents

Frequency of glucose-6-phosphate dehydrogenase deficiency in malaria patients from six African countries enrolled in two randomized anti-malarial clinical trials